When Is Short Selling Fraudulent?
Listen to the Audio Mises Wire version of this article.
Amid the controversy over GameStop, many cynics argued that something sinister was clearly afoot because the hedge funds had shorted 138 percent of the outstanding (floating) shares. In this article I’ll review that particular claim, as well as another seemingly dubious practice, so-called naked short selling. My conclusion is that shorting more than the total outstanding shares isn’t perverse or fraudulent, whereas naked short selling—depending on the context—might be.
A Review of Short Selling
Before diving into the specific variants, let me first explain the basics of short selling and why it can be a healthy activity in a free market. (In this section I reproduce material I published in an earlier mises.org article.)
In general, those who speculate in the stock market provide a “social” service—if they profit—by steering asset prices to their correct levels, as I explain in this article. If investors believe a particular stock is underpriced, they can buy shares of it and then unload them once the stock has met or surpassed what they view as its “proper” level. In this way, the speculators push up the (initially) underpriced stocks, helping to correct the “error” in the original price structure. Notice that it doesn't matter whether the investors who notice the initial underpricing own any of the stock at the outset.
However, things are different when an investor believes a stock is overpriced. If the investor happens to own shares of the stock in question, the obvious move is to sell some or all of the position, which both earns a relative gain for the investor and also speeds the downward move in price.
If this were the end of the story, there would be an obvious asymmetry in the market’s ability to rely on the dispersed knowledge of experts in diverse fields. Namely, it would mean that the only people who could act on their belief that a stock is overpriced would be those who already own the stock.
Fortunately, the market allows short selling, where someone who has zero shares of a stock can, in effect, sell shares and end up holding a negative position. In other words, a person “going long” might end up holding a hundred shares of a stock, whereas someone “going short” might end up holding minus a hundred shares.
Suppose stock XYZ is trading at $50, but Sam the Speculator believes tomorrow’s news will contain something very unfavorable for the stock. Further suppose that the rest of the market isn’t seeing things the way Sam is. Sam can borrow, say, a hundred shares of XYZ from Harold the Shareholder, sell them today for $5,000, and then wait for the news to hit. When it does, and the stock sinks to $40, Sam buys back the hundred shares for $4,000, and returns them to Harold (along with a fixed fee). Harold is better off, because he earned the fee; the price drop would have hit him in any case. (If Harold wants to unload his XYZ stock as the price drops but before Sam returns his shares, then Harold himself can short a hundred new shares and end up no worse off than he otherwise would have been.1) Meanwhile, Sam has netted $1,000 (minus the fee) from his superior foresight.
Things get more complicated when the short seller takes longer to return the shares; we have to worry about dividend payments, interest, etc. But the basic principle is simple enough: just as a speculator who wants to go long can borrow money to buy stocks, so too a speculator who wants to go short can borrow stocks to “buy money.” Short selling is no more mysterious than buying stocks on margin.
How Can There Be More Shorts Than Total Shares?
Now that we’ve reviewed the basic framework of short selling, we can answer the question: How can it be possible to short more shares than exist? The answer is pretty straightforward: the same share can be shorted multiple times as it passes hands from one owner to the next.
For example, suppose there are a total of a thousand shares of XYZ stock. Ten different speculators each want to short a hundred shares each. So each of the ten speculators makes arrangements with the original holders of XYZ stock in order to borrow a thousand total shares from them to sell at the original market price.
Now when a speculator shorts a stock, he is selling it. That means there must be a buyer of the shares on the other side of the transaction. When a new person enters the market to buy shares of a stock, there isn’t a Post-it note affixed to some of the shares saying “This share was obtained as part of a short sale, so it is forbidden to be lent out again.” No, all the shares are fungible, and each one just as “shortable” as any other. When a speculator borrows shares of XYZ in order to sell them (with the hopes that the price will go down), the person on the other side of the transaction is engaging in a normal purchase. That new owner acquires full rights to the shares, including the right to lend them out to a new short seller.
Therefore, continuing our scenario, suppose that after the original thousand shares of XYZ have been borrowed as part of ten different short sales the price of XYZ is still too high, in the view of a speculator. This speculator approaches the new owners of the stock, and borrows 380 shares from them to short. At this point, the total (gross) short position is 1,380 shares, even though there are only a thousand total shares in existence. In other words, the gross short position is 138 percent of the outstanding shares.
To understand the situation, consider the gross vs. net positions: there are currently a thousand shares owned outright by various individuals. There are also people who are owed 1,380 shares in total, because they lent out their shares for a short sale; they are “long” the XYZ stock, because they benefit if its price goes up. On the flip side of this, there are the short sellers who collectively owe 1,380 shares to the people from whom they borrowed the stock. (They are clearly “short” the stock, and benefit if its price drops.) Therefore, the gross short position is 1,380 shares, which is 138 percent of the total supply, but the gross long position is 2,380 shares. Thus, the net position is long 2,380 – 1,380 = 1,000 shares, which is 100 percent of the total quantity of shares. Nothing weird here, folks.
Before leaving this section, let me address one potential objection. Isn’t there a sense in which it’s physically impossible for the short sellers to close out their positions? After all, how can they collectively buy back 1,380 shares when there are only a thousand shares total?
The answer is that they don’t have to do this in one fell swoop. (This is also why it’s possible for a community to pay interest plus principal on money loans.) One of the original speculators can buy a hundred shares of XYZ and return them to the original owners, thus reducing the gross short position to 1,280 shares. Then a different speculator could buy another hundred shares and do the same, reducing the shorts to 1,180 shares, and so on. The cumulative short position can be unwound in stages, just as it was originally wound up in stages.
Now it’s true, the speculators might find it expensive to convince the current holders of the XYZ stock to sell some of their shares in order to close out the original short. But this is true regardless of the scale of the operation. In other words, whether the gross short position on XYZ is 138 percent or 1 percent, if the current holders hang on tight to their shares, it might be prohibitively expensive for the shorters to close out their positions. There is nothing peculiar to a gross short position being greater than 100 percent in this regard.
A More Dubious Practice: Naked Short Selling
With “naked” short selling, the transactions are marked up “on paper” without first locating and obtaining actual shares of the stock. To adapt our scenario, if a speculator wants to short a hundred shares of XYZ when it’s selling for $50, perhaps his brokerage will allow him to do so even though it hasn’t gotten possession of a hundred shares. When the speculator closes out his position, the brokerage calculates what his profit/loss would have been if he really had obtained the shares and credits/debits his account accordingly.
The potential problem here occurs if something goes wrong and the speculator can’t close out his original position because XYZ has gone up in price too much. For example, suppose that after shorting the hundred shares at $50 (and thus temporarily earning $5,000), the speculator is horrified to see that XYZ shoots up to $1,000 per share! In order to close out his position, the speculator would need to spend $1,000 x 100 = $100,000 to obtain the hundred shares on the open market and return them to the brokerage. If the speculator doesn’t have that kind of money, he can’t do it.
The victims in this type of scenario are the people who thought they were buying those hundred shares of XYZ back when it was selling for $50. (After all, that’s where the original $5,000 came from.) For small enough dollar amounts, the brokerage itself would cover the speculator’s loss. But in principle, if a brokerage allowed many of its clients to engage in naked short selling, then a surprising spike in stock prices could cause such aggregate losses that even the brokerage itself couldn't afford to cover them. In that case, investors who thought they were buying shares of stock from the brokerage would discover that even though it had taken their money and told them the transaction went through, they in fact did not own actual shares.
Readers will note the similarity between naked short selling and fractional reserve banking. It’s true in both cases that contracts could carefully spell out the details, and give the brokerage/bank the legal right to refuse redemption, but there is a line in Austrian economics (going from Mises through Rothbard to many of today’s Rothbardians) which argues that legal niceties don’t really matter, it’s commercial practice that does. If bank depositors treat demand deposits as immediately redeemable (and hence equivalent to money proper), then fractional reserve banking causes the boom-bust cycle. Likewise, one might argue as an Austrian that naked short selling is a dubious practice that could foster instability in the financial markets. (Note that my own position on this latter issue has evolved since I previously wrote on naked shorting for mises.org.)
In a free market, short selling is a healthy practice that allows farsighted speculators to push down overpriced stocks. There is nothing inherently dubious or fraudulent should the gross short position exceed the total number of shares. However, the practice of naked short selling could be problematic, if it became large relative to overall trading.
- 1. For example, suppose the bad news hits (as Sam anticipated but Harold did not), and the price of XYZ begins falling over the course of a few hours from $50 down to $40. If Harold originally would have closed out his position once XYZ fell to (say) $45, then he can still borrow and short a hundred shares once XYZ hits that threshold. Selling the borrowed shares at that price garners him $4,500, and then Harold gives the hundred shares plus the fee to the person from whom he borrowed the shares, once Sam closes out the original short position. When the dust settles, Harold ends up with $4,500, which is exactly what he would have had in an alternate universe where he never met Sam and instructed his broker to sell his shares of XYZ if they ever fell to $45.